INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

FORMATION OF NEW CAPILLARIES and thus the increase in the number of capillaries per muscle fiber occur under physiological circumstances during endurance training as part of the adaptive responses of the cardiovascular and skeletal muscle systems to improve O₂ transfer to myocytes (4, 5, 8). Angiogenesis can be considered as the culmination of several steps involving dissolution of the extracellular matrix underlying the endothelium, cell migration, and endothelial cell proliferation (21, 24). A number of naturally occurring growth factors and cytokines can induce and/or promote formation of new vessels by stimulating endothelial cell growth and migration. Growth factors known to be involved in such processes include vascular endothelial growth factor (VEGF) (27), basic fibroblast growth factor (bFGF) (17), and transforming growth factor-₁ (TGF-₁) (31), as well as several cytokines (34, 38).

It has been recently shown in intact rats (7, 19) that exercise induces greater expression of genes encoding angiogenic growth factors. Our laboratory demonstrated by quantitative Northern blot analyses (7) that the mRNA levels for VEGF in gastrocnemius muscle rose substantially immediately after a single bout of exercise in the untrained intact rat. The response was quantitatively related to exercise intensity and was augmented by hypoxia. The bFGF and TGF-₁ mRNA levels increased less with exercise, and hypoxia had no effect at all on bFGF message. Similarly, Hang et al. (19) demonstrated a significant increase of VEGF mRNA levels in anterior tibialis and extensor digitorum longus muscles in rats after electrical stimulation for up to 21 days. Interestingly, these authors (19) found higher VEGF mRNA levels in the early stages of muscle stimulation and gradually decreasing levels at later stages.

Many unanswered questions remain regarding the stimuli and underlying mechanisms that modulate transcription of angiogenic growth factors. It has been hypothesized that different biomechanical changes in the microcirculation linked with exercise, namely 1) shear stress within the vessels (21), 2) increased capillary wall tension (21, 22), and, 3) stretching of capillaries due to mechanical distortion (21, 22), may play a significant role in the upregulation of angiogenic growth factors. It is well accepted that increased blood flow increases the velocity of flow in capillaries (40) and thus calculated shear stress within vessels (11, 20). A competing hypothesis (1) is that increased metabolic requirements of the tissue cells is the primary factor leading to angiogenesis. In this hypothesis, O₂ has been implicated as a major control element, because vessel growth increases during hypoxic conditions (1, 34, 39) and decreases during hyperoxic conditions. Intracellular PO₂ falls to very low levels (3 Torr) in normal human muscle during exercise (37).

In the present investigation, a canine preparation was used to determine whether acute passive hyperperfusion of one leg, presumably via biomechanical changes in the microcirculation, induces a significant increase in mRNA abundance for VEGF, bFGF, and TGF-₁. The study was designed to compare the effects of passive hyperperfusion alone on angiogenic growth factor mRNA responses vs. those provoked by contraction-induced hyperperfusion. In both instances, changes in gene expression of the test leg were compared with those of the corresponding resting muscle groups of the contralateral leg, which served as controls, pump perfused at intact resting blood flow rates.

	METHODS

Top Abstract Introduction Methods Results Discussion References

This study was approved by the University of California, San Diego, Animal Subjects Committee. Ten adult mongrel dogs of either sex, weighing 17.7 ± 1.8 kg, were initially studied, but three of them were excluded from the study due to technical problems reported in Experimental protocol. The animals were anesthetized with 30 mg/kg of pentobarbital sodium, and maintenance doses were given as required. The dogs were intubated with an endotracheal tube, the cuff of which could be inflated to ensure a tight seal. Heating pads were used to maintain esophageal temperature at 37°C. Ventilation with room air was maintained by using a tidal volume of ~15 ml/kg at a respiratory rate required to preserve normal values of arterial PO₂ and PCO₂ throughout the study with the use of a Harvard 613 ventilator. End-expired O₂ and CO₂ and airway pressure were monitored during the study. Positive end-expiratory pressure of ~5 cmH₂O was applied throughout the study. The animals were given heparin at a dosage of 1,500 U/kg after surgical preparation.

Dog preparation. One carotid artery was cannulated to monitor systemic blood pressure and for arterial blood sampling. The proximal portion of the femoral artery just below the inguinal ligament of both legs (test leg and control leg) was exposed over a length of ~7 cm, and resting blood flow was measured (see Experimental protocol). The two ends of a single, heparinized, saline-filled length of plastic tubing were introduced into the femoral artery proximally and distally, respectively. During the surgical procedure, special care was taken to prevent ties of collateral branches of the femoral artery that might perturb blood flow to the skeletal muscles in the hindlimb of the dog. The outflow from the proximal end of the femoral artery was passed through this tubing via a roller pump (Sigmamotor, Cole-Palmer Instrument, Chicago, IL) for controlled perfusion to the distal portion of the femoral artery. A pressure transducer in the distal part of the arterial line constantly monitored perfusion pressure. An electromagnetic flow probe (T106-3RB, Transonic Systems, Ithaca, NY) was placed distally to the pump around the femoral artery to monitor blood flow in the test leg throughout the study. In those animals in which electrical stimulation was performed (protocol B), after initial measurements of resting blood flow of both legs, the femoral and sciatic nerves of the test leg were doubly ligated and cut between ties.

Experimental protocol. Protocol A (passive hyperperfusion) initially included five dogs, but data from two of them could not be used. Dog 1 was excluded before any molecular analysis was performed, because the quality of the surgical procedure and maintenance of the animal preparation during the study were considered suboptimal (hypovolemic shock), and dog 4 could not be used because of mRNA degradation in all muscle biopsies taken. Thus only the three remaining dogs (dogs 2, 3, and 5) are reported in this protocol. A total of 18 individual muscles, representing 10 different muscle groups from the 3 dogs, was sampled, and mRNA levels were measured.

mRNA isolation and Northern analysis. Total cellular RNA was isolated from each muscle sample by the method of Chomczynski and Sacchi (10). RNA preparations were quantitated by absorbance at 260 nm, and intactness was assessed by ethidium bromide staining. Ten micrograms of total cellular RNA were separated by electrophoresis in 6.6% formaldehyde-1% agarose gel. Fractionated RNA was transferred by Northern blot to a Zeta probe membrane (Bio-Rad, Hercules, CA). RNA was cross-linked to the membrane by ultraviolet irradiation by using an ultraviolet cross-linker (model FD-UVXL 1000, Fisher Scientific) and stored at 4°C. The blots were then probed with oligolabeled [-³²P]deoxycytidine triphosphate cDNA probes, which had a specific activity of 1 × 10⁹ disintegrations · min¹ · µg DNA¹ (13). The human VEGF probe is a 0.93-kb cDNA fragment isolated from the EcoR I site of pUC-derived plasmid (27). The human TGF-₁ cDNA probe is a 0.985-kb Hind III/Xba I insert cloned into pBluescript II KS⁺ vector (36). The bFGF probe is a 1-kb Xho I fragment of human bFGF cDNA (24). Prehybridization and hybridizations were performed in 50% formamide, 5× SSC (20× SSC is 0.3 M sodium chloride and 0.3 M sodium citrate), 10× Denhardt's solution (100× Denhardt's solution is 2% Ficoll and 2% polyvinyl pyrrolidone), 50 mM sodium phosphate (pH 6.5), 1% SDS, and 250 µg/ml of sonicated salmon sperm DNA at 37 or 42°C. Blots were washed with 2× SSC and 0.1% SDS at room temperature, with 0.1× SSC and 0.1% SDS at 55°C for the VEGF mRNA, and with 1× SSC and 0.1% SDS at 60°C for TGF-₁ and bFGF mRNAs. Blots were exposed to XAR-5 X-ray film (Eastman Kodak, New Haven, CT) by use of a Cronex Lightning Plus screen at 70°C. In all analyses, quantitative densitometry of the autoradiographs was used to measure the mRNA levels for all three growth factors. Each blot was subsequently reprobed (after prior complexes were stripped) with a cDNA specific for 18S ribosomal RNA, and this signal was used to normalize the mRNA signal for minor variations in the lane loading. Because acceptable reproducibility of the technique in our hands has been previously demonstrated (7), duplicate analyses were not performed on the samples.

Statistical treatment. For the three growth factors, results were examined as the ratio of 18S-normalized mRNA levels in the test leg compared with those observed in the same muscle of the contralateral (control) leg of each animal, such that lack of effect of any experimental condition would give rise to a ratio of 1.0. This analysis was independently performed for the two different protocols: protocol A, passive hyperperfusion only (3 dogs, n = 18 samples); and protocol B, muscle contraction by electrical stimulation (4 dogs, n = 10 samples). In each protocol (passive hyperperfusion and contraction-induced hyperperfusion), 12 muscle samples were not considered in the analyses because of mRNA degradation in either the experimental or the control leg. The results represent an average of the effects of each experimental condition in all individual muscles sampled. The null hypothesis was tested for each condition by using Student's paired t-test comparing the ratio to 1.0. Comparisons between groups of animals were done by using an unpaired analysis. Statistical significance was accepted if P < 0.05. Results are expressed as means ± SE.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

In protocol A (passive hyperperfusion), blood flow in the test leg was 5.4 ± 0.87-fold (range 3.8-6.8) higher than under control conditions. No statistically significant differences were observed in blood flow measured during electrical stimulation [5.9 ± 0.60-fold (range 4.7-6.5)]. Although the relative increases in flow were not different between the two protocols, it is possible that differences in vascular resistance could have exposed the muscle microvasculature to different hemodynamic conditions, which could then be argued to explain the VEGF mRNA results despite similar increases in flow per se. To examine this possibility, the increase in flow divided by the increase in perfusion pressure (from rest to high-flow state) was computed for each dog and compared. These ratios were not significantly different (P = 0.4) at 2.7 l · min¹ · mmHg¹ in the passively perfused group and 3.5 l · min¹ · mmHg¹ in the electrically stimulated muscles group, suggesting that the microvascular beds of the two groups were reflecting similar hemodynamic conditions. Averaged for all dogs, absolute values for resting and high blood flows in the test leg were 58 ± 7.3 and 320 ± 37.6 ml/min, respectively.

Apparently negative results always require special effort to minimize the possibility that a type II (false negative) error has occurred. Consequently, we reanalyzed the data 1) without the 18S RNA normalization, because in some cases the densitometer may have been reading beyond its linear range, and 2) by nonparametric tests (Wilcoxon's signed rank test). The outcome in both cases was the same as originally seen: passive hyperperfusion did not alter VEGF mRNA, whereas electrically stimulated contraction did.

Figures 1 and 2 show the Northern blots from protocol A (high blood flow by passive hyperperfusion). It is clear that VEGF (Fig. 1) and bFGF mRNA levels (Fig. 2) did not show significant changes immediately after high blood flow for 1 h. As shown in the figure, TGF-₁ slightly but significantly increased. Figures 3 and 4 portray the Northern blots in protocol B (similar high flow during electrical stimulation of the femoral and sciatic nerves). An increase in the abundance of VEGF mRNA (Fig. 3) can be seen in the contracting muscle groups (gastrocnemius, sartorius, vastus medialis, rectus femoris, and vastus lateralis) indicated in the figure, but significant changes were not observed in bFGF or TGF-₁ (Fig. 4).

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Northern blots of mRNA signals for vascular endothelial growth factor (VEGF) in different muscle groups of 3 dogs (dogs 2, 3, 5) included in protocol A (passive hyperperfusion). Total cellular RNA (10 µg) was loaded in each lane. Blots were reprobed and normalized to 18S ribosomal RNA (18S) signal. Low blood flow (L), resting blood flow in control leg; high blood flow (H), high-flow condition in test leg; CT, cranial tibial; G, gastrocnemius; GR, gracilis; SM, semimembranosus; B, biceps femoris; S, sartorius; ST, semitendinosus; VM, vastus medialis; RF, rectus femoris. To produce figures, lanes corresponding to samples with degraded RNA were cut from the illustration, and autoradiograms from different blots were used and pieced together in a final composite.

View larger version (43K): [in this window] [in a new window]   Fig. 2.   Northern blots of mRNA signals for basic fibroblast growth factor (bFGF) and transforming growth factor-1 (TGF-1) in different muscle groups of 3 dogs (dogs 2, 3, 5) included in protocol A (passive hyperperfusion). Blots were reprobed and normalized to 18S ribosomal RNA signal.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Northern blots of mRNA signals for VEGF in different muscle groups of 4 dogs (dogs 7, 8, 9, 10) included in protocol B (electrical stimulation). Total cellular RNA (10 µg) was loaded in each lane. Blots were reprobed and normalized at 18S ribosomal RNA signal. Rest (R), resting blood flow in control leg; exercise (E), high-flow condition in test leg after electrical stimulation; VL, vastus lateralis.

View larger version (31K): [in this window] [in a new window]   Fig. 4.   Northern blots of mRNA signals for bFGF and TGF-1 in different muscle groups of 4 dogs (dogs 7, 8, 9, 10) included in protocol B (electrical stimulation). Lane loading is controlled by 18S ribosomal RNA.

Figure 5 portrays the quantitative densitometry (mRNA levels normalized by 18S ribosomal RNA) for the three growth factors: VEGF, bFGF, and TGF-_1. Results (y-axes) are expressed as the high flow-to-rest flow ratio for each experimental condition. In the three panels, the left bar indicates the change in mRNA expression during passive hyperperfusion alone (protocol A), and the right bar corresponds to the signal obtained at similar high-flow levels in the electrically stimulated muscles (protocol B). As indicated above, we observed no significant increase in the mRNA expression of two of the growth factors (VEGF and bFGF) during passive hyperperfusion: however, TGF-₁ was only slightly elevated (14%; P < 0.03). In contrast, more than a threefold increase in VEGF mRNA abundance (P < 0.02) was observed after electrical stimulation. There were no significant changes in both bFGF and TGF-₁ mRNA levels in this experimental condition.

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Quantitative densitometry of Northern blots of mRNA signals for VEGF (top), bFGF (middle), and TGF-1 (bottom). For the 3 growth factors, results are expressed as ratios of mRNA levels in test leg to those observed in corresponding muscle group of control leg, such that lack of effect of any experimental condition would give rise to a ratio of 1 (dashed lines). Left bars correspond to results for protocol A (passive hyperperfusion), and right bars to protocol B (electrical stimulation). Bars represent means (± SE) of ratios of experimental leg to control leg for all individual muscle groups sampled and analyzed. NS, not significant.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The study by Breen et al. (7) showed that mRNA levels of angiogenic growth factors (VEGF, bFGF, and TGF-₁) in rat gastrocnemius increased significantly after 1 h of normoxic exercise in the intact animals. The response was quantitatively related to the exercise intensity and was augmented when the rats were exercised while breathing a low inspired O₂ concentration of 0.12. These results (7), as well as those reported by Hang et al. (19) using electrical stimulation, show that muscle activity leads to an early increase of mRNA levels for angiogenic growth factors. Local hypoxia (1, 34, 39) and several other stimuli, some of them potentially related to inadequate tissue cell oxygenation, such as changes in local pH, release of microvascular vasodilators, or modifications of local concentrations of metabolites, may be postulated as triggers of angiogenesis induced by exercise.

A competing hypothesis is that biomechanical events at the level of the microcirculation (shear stress within vessels, capillary wall tension, or vascular stretch; Refs. 21, 22) may play a significant role in exercise-induced angiogenesis. Growth of new vessels induced after long-term administration of different types of vasodilators (21) has been attributed to biomechanical effects of increased blood flow through increase in shear stress and/or capillary wall tension. It has been shown that local release of polypeptides that modulate smooth muscle growth is signaled by changes in the shape of endothelial cells (15, 23), from a spheroid to a flat configuration, and also by the reduction of the contacts between adjacent endothelial cells (9, 26). The angiogenic role of biomechanical forces (20, 33) is substantially reinforced by the observation that capillary sprouts in chronically stimulated skeletal muscle are found predominantly at sites of maximum curvature of preexisting capillaries.

However, the main finding in the present study is that passive hyperperfusion (5.4-fold increase in blood flow alone) maintained for 1 h did not produce significant changes in mRNA levels for VEGF and bFGF. The change in TGF-₁, while statistically significant, was relatively small. In contrast, a similar increase in blood flow occurring during electrical stimulation provoked a more than threefold increase in VEGF mRNA levels (Fig. 5). These results therefore suggest that, if biomechanical changes in the skeletal muscle microcirculation occur with fivefold increases in blood flow induced by exercise, they do not play a significant role in the mRNA abundance of angiogenic growth factors. In contrast, short-term increases in metabolic muscle requirements, provoked by short-term exercise in intact animals (7) and during short- (present study) or long-term electrical stimulation (19), induce a significant mRNA response of angiogenic growth factors, mainly for VEGF. It must be pointed out, however, that the present study does not address whether increased VEGF mRNA abundance represents increased transcription and/or increased message stability within the myocyte.

A potential limitation of the present study is that the lack of change in the average mRNA responses of angiogenic growth factors after passive hyperperfusion (Fig. 1) could reflect the balance between muscles that were highly perfused and increased in mRNA VEGF levels and those that were less well perfused through nonuniform flow distribution such that mRNA levels went down. We argue that such a possibility was ruled out by the uniformity of the response to high blood flow over all 18 muscles (Figs. 1 and 2). The low variance in the response to passive hyperperfusion is illustrated by the low standard errors of the mRNA ratios for VEGF, bFGF, and TGF-₁ depicted in Fig. 5, left bars. Moreover, as total femoral flow was passively increased over fivefold, it is hard to imagine that, in any muscle, flow under these conditions could be less than when femoral flow was at fivefold lower resting levels.

After electrical stimulation (Fig. 5, right bars), the dissociation between the increase in VEGF mRNA levels and the lack of changes in mRNA levels for bFGF and TGF-₁ is not unexpected. The study by Breen et al. (7) also showed a higher signal for VEGF than for bFGF and TGF-₁. In addition, the four 3-min electrical stimulation periods may not produce the same effects as natural exercise for 1 h, and one must also remember that the data of Breen et al. were for another species (rat). We should take into account that the results may also be related to the time of sampling relative to the onset of the stimulus (7, 39). Possibly, passive hyperperfusion could affect growth factor mRNA levels, but we failed to see this because of our sampling protocol, i.e., taking the muscle for analysis at a single time point within minutes of the end of contractions or hyperperfusion. However, because the earlier work from our laboratory showed immediate increases in mRNA signals (7) that were reproduced in the contracting muscles from the present study, we would argue that, even if we missed seeing some delayed response, any such response would not reflect the exercise-induced response of the intact muscle that we were attempting to explain in the first place.

VEGF, also called vascular permeability factor, is actually a family of several highly specific, potent growth factors inducing endothelial mitogenesis (25, 35). The key importance of VEGF in initial vascular development is emphasized by the fact that removal of the VEGF gene is lethal early in embryogenesis in animals heterozygous for gene mutation (14) and that VEGF inactivation reduces angiogenesis in tumors (6, 41) and in the eye (2, 3). Upregulation of VEGF gene expression may represent an early response to increased metabolic requirements that corresponds with the short half-time of VEGF mRNA levels determined in vascular smooth muscle cells (29). bFGF is also a direct angiogenic factor; however, it has a broader cell-type specificity. It has been suggested that bFGF may represent a more long-term response to muscle stimulation (31). TGF-₁ is an indirect angiogenic factor with a wide variety of effects on cell proliferation and regulation of extracellular matrix components.

The dissociation between the response of VEGF to electrical stimulation and the lack of signal in both bFGF and TGF-₁ may also suggest differences in promoter regions in the genes of these angiogenic growth factors. Whereas hypoxic regulatory sequences in both the 5' and 3' regulatory regions of the VEGF gene (16, 28, 30) have been clearly demonstrated, this is not the case for bFGF and TGF-₁. It is also possible that TGF-₁ and bFGF are activated by release of mediators from the cell or extracellular matrix (12, 32) and not stimulated at the transcriptional level. Furthermore, both bFGF and TGF-₁ seem to be involved in the regulation of VEGF, resulting in a network of growth factor interactions. Although the above interpretation is still partly speculative, the dissociation of the VEGF mRNA response to electrical stimulation provides a rationale for further search of the regulatory regions in the genes of these angiogenic growth factors. Recently, Gu and Adair (18) showed that hypoxia can induce the expression of VEGF mRNA and its protein in cultured dog myocardial vascular smooth muscle cells, suggesting that vascular smooth muscle cells might also play an angiogenic role in the regulation of the expression of VEGF.

In summary, the present study demonstrates that 1 h of passive hyperperfusion alone does not induce increases in mRNA levels for VEGF or bFGF but does induce minimal increases (14%) in those for TGF-₁. In contrast, VEGF mRNA levels show a marked enhancement after muscle contraction at similar high blood flow. Our results support the hypothesis that early exercise-induced angiogenesis is primarily related to high cellular energy requirements rather than to biomechanical changes in the microcirculation secondary to increased blood flow. Accordingly, we speculate that intracellular hypoxia, metabolite concentration changes, or mechanical effects of muscle contraction might play more important roles in the early upregulation of the expression of the VEGF gene during exercise. In light of the present study, angiogenesis induced by microvascular biomechanical events may require prolonged exposure to the accompanying hemodynamic changes.